Definition Source
This dictionary defines the core metrics used by GEO · Compare2Best when discussing B2B AI search visibility. The purpose is to make GEO measurement auditable: every metric should connect back to query sets, captured answers, citation evidence, and known limitations.
| Metric | Definition | Interpretation boundary |
|---|---|---|
| AI Visibility Index | A composite view of whether a brand appears in AI-generated answers for a defined buyer query set. It should be tied to exact prompts, platforms, timestamps, and evidence links. | Use only with a fixed query set and documented capture method. |
| Brand Mention Rate | The percentage of tested AI answers where the brand is named, whether cited or uncited. | Use only with a fixed query set and documented capture method. |
| Citation Rate | The percentage of tested AI answers that include an inspectable source link or source reference supporting the brand, page, or claim. | Use only with a fixed query set and documented capture method. |
| Citation Share | A brand’s share of total cited sources within a defined category query set. | Use only with a fixed query set and documented capture method. |
| Citation Stability | How consistently the same brand or source appears across repeated tests over time. | Use only with a fixed query set and documented capture method. |
| Competitive Displacement | A case where competitors are recommended or cited for buyer queries where the monitored brand is absent. | Use only with a fixed query set and documented capture method. |
| Fact Accuracy Rate | The percentage of AI claims about a brand that match verified company facts, certifications, products, and capabilities. | Use only with a fixed query set and documented capture method. |
| Entity Completeness | The degree to which the brand’s organization, products, certifications, addresses, authors, and content assets are machine-readable and internally consistent. | Use only with a fixed query set and documented capture method. |
| Answer Source Diversity | The number and quality of independent source types supporting a brand in AI answers: owned site, industry media, standards bodies, directories, reports, and community references. | Use only with a fixed query set and documented capture method. |
| AAOV Score | AI Authority & Optimization Value: a composite diagnostic score combining current AI authority evidence and the size of the practical optimization gap. It is a planning model, not a guaranteed ranking score. | Use only with a fixed query set and documented capture method. |
| Hallucination Risk | The probability that AI systems describe a company with outdated, unsupported, or false claims. | Use only with a fixed query set and documented capture method. |
| Procurement-Intent Query | A buyer query that indicates sourcing, verification, comparison, certification, risk control, or supplier shortlisting intent. | Use only with a fixed query set and documented capture method. |
How AAOV should be used
AAOV is useful for prioritization, not for making guaranteed performance claims. A high AAOV opportunity means the brand has either visible authority gaps, weak entity evidence, missing structured data, or competitor displacement in procurement-intent answers.
Minimum evidence required for a GEO metric
- Exact query text and buyer intent class.
- AI platform, language, date, and region if available.
- Captured answer text and cited URL list.
- Manual review label: accurate, unsupported, outdated, hallucinated, or competitor-only.
- Repeated tests when monitoring trend change.
No Score Without Evidence
A dashboard score without prompt logs, source URLs, and manual verification is not decision-grade GEO evidence. It may be useful as a rough signal, but it should not be sold as proof of AI visibility.